Leaderboard Explore Compare Profile Community Blog FAQ Competition 2024 Report Issue

Sample stimuli

How to use

from brainscore_vision import load_benchmark
benchmark = load_benchmark("Kato2026Dense-error_consistency")
score = benchmark(my_model)

Benchmark API

Code examples

Model scores

Score Legend Min Alignment Max Alignment

Rank	Model	Score
1	vit_huge_clip336_laion	.541
2	convnext_xxlarge_clip	.528
3	dinov2_vitl14_lc_res336	.527
4	convnext_xxlarge:clip_laion2b_soup_ft_in1k	.504
5	dinov2_vitl14_lc	.504
6	dinov2_vitl14_reg_lc	.496
7	vit_huge_patch14_clip_336:laion2b_ft_in12k_in1k	.493
8	dinov2_vitg14_lc	.485
9	dinov2_vitg14_reg_lc	.470
10	vit_huge_patch14_clip_224:laion2b_ft_in12k_in1k	.412
11	vit_large_clip336_openai	.405
12	convnext_large_clip_laion2b_finelayers	.388
13	vit_large_patch14_clip_224:openai_ft_in12k_in1k	.386
14	AdvProp_efficientnet-b7	.374
15	vit_large_patch14_clip_336:openai_ft_in12k_in1k	.368
16	eva02_large_in22k	.323
17	vit_large_patch14_clip_224:openai_ft_in1k	.303
18	resnext101_32x32d_wsl	.210
19	convnext_large_mlp:clip_laion2b_augreg_ft_in1k_384	.208
20	resnext101_32x48d_wsl	.179
21	convnext_xlarge:fb_in22k_ft_in1k	.167
22	AdvProp_efficientnet-b6	.163
23	resnext101_32x16d_wsl	.157
24	vit_large_patch14_clip_224:laion2b_ft_in1k	.152
25	vit_large_patch14_clip_224:laion2b_ft_in12k_in1k	.149
26	resnet50-SIN	.126
27	AdvProp_efficientnet-b4	.125
28	deit3_base_patch16_224_fb_in22k_ft_in1k_gram_aligned_dinov3	.112
29	vit_large_patch14_clip_336:laion2b_ft_in1k	.090
30	ReAlnet04	.086
31	convnext_large:fb_in22k_ft_in1k	.084
32	cvt_cvt-13-384-in22k_finetuned-in1k_4_LucyV4	.083
33	resnet50-moclr8deg	.081
34	resnet_50_v1_spiking_l4a	.078
35	resnet50_byol_linhead	.076
36	swin_small_patch4_window7_224:ms_in22k_ft_in1k	.070
37	densenet-201	.069
38	resnet50_barlowtwins_linhead	.063
39	ReAlnet10	.061
40	ReAlnet08	.059
41	vit_base_patch16_clip_224:openai_ft_in12k_in1k	.058
42	antialias-resnet152	.055
43	ReAlnet09	.054
44	resnet34_imagenet_full	.054
45	convnext_large_in22k_supervised	.054
46	resnet50_vicreg_linhead	.051
47	regnety_032	.051
48	cvt_cvt-w24-384-in22k_finetuned-in1k_4	.048
49	deit3_base_patch16_224_fb_in22k_ft_in1k	.048
50	convnext_base:clip_laiona_augreg_ft_in1k_384	.045
51	grcnn_109	.044
52	ReAlnet07	.044
53	resnet50_robust_l2_eps3	.043
54	imagenet_l2_3_0	.043
55	resnet-50-robust	.043
56	resnet_101_v1	.041
57	eicircuit_resnet18_bn_20260919	.038
58	resnet50-SIN_IN	.038
59	focalnet_tiny_lrf_in1k	.034
60	vit_base_patch16_clip_224:openai_ft_in1k	.031
61	resnet50_ewconplus_wzc_seed2_2__323	.030
62	vit_relpos_base_patch16_clsgap_224:sw_in1k	.029
63	resnet50_hvs_wzc	.029
64	deit_large_imagenet_full_seed-0	.028
65	convnext_base_imagenet_full_seed-0	.027
66	AdvProp_efficientnet-b8	.026
67	resnet_101_v2	.026
68	convnext_large_imagenet_full_seed-0	.026
69	AT_efficientnet-b2	.023
70	AdvProp_efficientnet-b2	.023
71	deit3_base_patch16_224_fb_in1k	.021
72	convnext_tiny_sup	.019
73	mobilenet_v2_1_4_224	.014
74	resnet50_st_wzc_seed3_2_323	.014
75	resnet50_simclr_linhead_01norm	.013
76	densenet-121	.012
77	cvt_cvt-21-384-in22k_finetuned-in1k_4_LucyV4	.011
78	resnet_50_v1	.011
79	resnet50_tutorial	.011
80	resnet50-sup	.011
81	convnext_base_fb_in1k	.010
82	vit_relpos_base_patch32_plus_rpn_256:sw_in1k	.009
83	convnext_base_fb_in1k_gram_aligned_dinov3	.008
84	resnext101_32x8d_wsl	.006
85	resnet50_robust_l2_eps1	.006
86	convnext_base_fb_in22k_ft_in1k	.005
87	shufflenet_v2_x1_0	.004
88	mehrer2020_alexnet_training_seed_01_fov16	.001
89	pnasnet_large_halfPrecision	.000
90	grcnn	.000
91	mehrer2020_alexnet_training_seed_05	.000
92	resnet_50_v2	.000
93	AlexNet_SIN_fov12	.000
94	convnext_femto_ols:d1_in1k	.000
95	resnet50-vitoimagevidnet8	.000
96	resnet50-VITO-8deg-cc	.000
97	alexnet	.000
98	AlexNet_SIN	.000
99	vit_tiny_r_s16_p8_384:augreg_in21k_ft_in1k	.000
100	resnet50_st_wzc_2	.000
101	cvt_cvt-13-224-in1k_4_LucyV4	.000
102	convnext_small_imagenet_100_seed-0	.000
103	effnetb1_cutmixpatch_augmix_robust32_avge4e7_manylayers_324x288	.000
104	cv_18_dagger_408_pretrained	.000
105	convnext_tiny_imagenet_full_seed-0	.000
106	efficientnet_b2_imagenet_full	.000
107	resnet101_imagenet_full	.000
108	effnetb1_272x240	.000
109	inception_v3	.000
110	resnet50_ewconplus_wzc_seed6_2_323	.000
111	convnext_base_fb_in22k_ft_in1k_gram_aligned_dinov3	.000
112	deit_small_imagenet_full_seed-0	.000
113	mehrer2020_alexnet_training_seed_03	.000
114	efficientnet_b0	.000
115	resnet50_st_wzc_seed7_2_323	.000
116	effnetb1_cutmix_augmix_sam_e1_5avg_424x377	.000
117	antialiased-r50	.000
118	voneresnet-50-robust	.000
119	deit_base_imagenet_full_seed-0	.000
120	resnet_50_v1_spiking	.000
121	cvt_cvt-13-384-in1k_4_LucyV4	.000
122	vonegrcnn_62e_nobn	.000
123	resnet_152_v2	.000
124	efficientnet_b0_imagenet_full	.000
125	cvt_cvt-21-384-in1k_4_LucyV4	.000
126	ijepa_vith14_linhead	.000
127	convnext_small_imagenet_full_seed-0	.000
128	resnet50_simsiam_linhead	.000
129	resnet_SIN_IN_FT_IN	.000
130	resnet50-SIN_IN_IN	.000
131	vonegrcnn_47e	.000
132	cvt_cvt-21-224-in1k_4_LucyV4	.000
133	resnet50_st_wzc_seed6_2_323	.000
134	effnetb1_cutmixpatch_SAM_robust32_avge6e8e9e10_manylayers_324x288	.000
135	deit3_base_patch16_224_fb_in1k_gram_aligned_dinov3	.000
136	custom_model_cv_18_dagger_408	.000
137	efficientnet_b1_imagenet_full	.000
138	resnet152_imagenet_full	.000
139	convnext_tiny:in12k_ft_in1k	.000
140	resnet50_ewconplus_wzc	.000
141	mehrer2020_alexnet_training_seed_02	.000
142	resnet50_imagenet_full	.000
143	mehrer2020_alexnet_training_seed_01_fov12	.000
144	ViT_L_32_imagenet1k	.000
145	Res2Net50_26w_4s	.000
146	resnet50_ewconplus_wzc_seed5_2__323	.000
147	resnet18_imagenet_full	.000
148	voneresnet-50-non_stochastic	.000
149	seresnext50_32x4d	.000
150	mehrer2020_alexnet_training_seed_04	.000
151	mehrer2020_alexnet_training_seed_01	.000
152	voneresnet-50	.000
153	AlexNet_SIN_fov16	.000
154	vgg_16	.000
155	nasnet_mobile	.000
156	blt_vs	.000
157	mehrer2020_alexnet_training_seed_01_fov4
158	mobilenet_v2_1_0_128
159	dinov2_vitl14_reg_lc_featread
160	gabor_filter_entropy_4x4_v2_forcedCosine
161	gabor_filter_entropy_4x4_v2
162	deepeisnn_ei_imagenet_resnet18
163	alexnet_training_seed_02
164	alexnet_training_seed_10
165	alexnet_training_seed_09
166	alexnet_training_seed_07
167	alexnet_training_seed_06
168	alexnet_training_seed_03
169	alexnet_training_seed_01
170	hmax
171	ReAlnet01_cornet_fov4
172	VOneCORnet-S
173	ReAlnet01_cornet
174	compact_resnet50_robust_V4
175	ReAlnet04_cornet
176	ReAlnet03_cornet

Benchmark bibtex

None

Ceiling

0.36.

Note that scores are relative to this ceiling.

Data: Kato2026Dense

Metric: error_consistency